t8star/minimax_h3_ref2va_patchin_hf102

MiniMax H3 Ref2VA Patch-In HF 1.02

Experimental ComfyUI single-file checkpoint derived from MiniMaxAI/MiniMax-H3 Ref2VA and the Comfy-Org/MiniMax-H3 INT8 ConvRot repack.

This is an experimental weight modification, not an official MiniMax release and not a proven “de-oil”, “de-wax”, restoration, or quality-fix model. The original checkpoint remains the recommended default.

MiniMax H3 is governed by the MiniMax H3 Community License Agreement, which defines excluded territories and mandatory redistribution conditions. Before publishing or redistributing this derivative, include the official LICENSE, keep the modification notice and NOTICE, and confirm that the intended distribution method and audience are authorized. A public Hugging Face repository may be reachable from excluded territories; a repository gate alone is not necessarily geographic access control. This model card is not legal advice.

中文说明

这是 MiniMax H3 Ref2VA 的实验性 ComfyUI 单文件衍生模型。它没有训练、微调或蒸馏，只对 视频 patch 输入投影中的 2×2 空间高频分量增加 2%。两组固定条件测试都出现了很弱的皮肤 高频代理正增益，但肉眼仍未确认能够消除 Ref2VA 的油感或蜡感，所以只能作为 EXP 对照模型。

发布前必须阅读上游许可证。该许可证对适用地域、公开分发、商业使用、安全措施、修改声明、 LICENSE 和 NOTICE 都有要求。

Model description

Property	Value
Task family	MiniMax H3 Base Ref2VA / reference-to-audio-video
Format	ComfyUI single-file diffusion model
Precision	Native INT8 ConvRot container inherited from the source repack
Training	None
Modified tensor	video_patch_proj.weight only
Input patch HF gain	1.02
Output-head gain	1.00 — unchanged
File size	34,038,894,550 bytes, approximately 31.70 GiB
Status	Experimental A/B candidate

What was changed

MiniMax H3 video latents are patchified with a 1 × 2 × 2 patch. For every latent channel, the four spatial input columns of video_patch_proj.weight were transformed in an orthonormal 2×2 Haar basis:

• the DC/common component is preserved at gain 1.00;
• the three non-DC H/V/D components are multiplied by 1.02;
• the result is transformed back and written into the input projection;
• no output head, audio tensor, shared Transformer block, VAE, text encoder, or ComfyUI node was changed.

The equivalent 4×4 transform has diagonal 1.015 and off-diagonal -0.005. Its all-ones/DC eigenvector has gain 1.00; the three orthogonal spatial-detail eigenvectors have gain 1.02.

This is not an output sharpening filter. It changes the video latent input projection used during each joint audio-video denoising forward pass. It cannot reconstruct real texture that the model does not generate.

Checkpoint integrity

Source

File: minimax_h3_ref2va_int8_convrot.safetensors
Repository: Comfy-Org/MiniMax-H3
Size: 34,038,894,550 bytes
SHA-256: 9eef934046a0671bc8a5daf87100705e1478419c574cfde70c50fbe6885f76a9

The source size and SHA-256 match the LFS object published by Comfy-Org/MiniMax-H3.

Modified checkpoint

File: minimax_h3_ref2va_patchin_hf102.safetensors
Size: 34,038,894,550 bytes
SHA-256: 5552621452ff73c9627ece6bd08f08e8c08b6504ee8a753b1aa1e1abcb70d9e4

Structural verification:

• safetensors header identical to the source: yes;
• tensor keys and shapes identical to the source: yes;
• changed tensor: video_patch_proj.weight only;
• changed byte interval: [17,879,061,191, 17,881,125,575);
• target writeback maximum absolute error: 0;
• DC-sum maximum absolute error: 9.5367431640625e-07;
• SHA-256 of all unchanged file regions: 82d4866debf06c9b9b7a67d645d1dcac57bdd91eba3b341a55a4e1fe9f687852.

The machine-readable surgery record is available at evidence/surgery.json.

ComfyUI usage

Required model components

This repository contains only the modified Ref2VA diffusion model. Obtain the remaining H3 components from the official or ComfyUI repack repositories under their applicable licenses:

• MiniMax H3-compatible Qwen3-VL text encoder;
• minimax_h3_video_vae_fp16.safetensors;
• minimax_h3_audio_vae_fp32.safetensors;
• a current ComfyUI build with MiniMax H3 and native INT8 ConvRot support.

Suggested layout:

ComfyUI/
└── models/
    ├── diffusion_models/
    │   └── minimax_h3_ref2va_patchin_hf102.safetensors
    ├── text_encoders/
    │   └── qwen3vl_32b_minimax_h3_*.safetensors
    └── vae/
        ├── minimax_h3_video_vae_fp16.safetensors
        └── minimax_h3_audio_vae_fp32.safetensors

Load the checkpoint with ComfyUI's Load Diffusion Model node and use a Ref2VA/R2V workflow. The official ComfyUI template is available at video_minimax_h3_r2v.json.

Recommended comparison protocol

For a meaningful A/B, change only the diffusion model:

1. generate once with minimax_h3_ref2va_int8_convrot.safetensors;
2. generate again with minimax_h3_ref2va_patchin_hf102.safetensors;
3. keep seed, reference media, prompt, dimensions, frame count, sampler, scheduler, shifts, steps, VAEs, text encoder and conditioning identical;
4. do not add a Turbo/four-step LoRA when reproducing the validation below;
5. inspect skin, hair, fabric, temporal consistency, identity, composition and audio together.

The local validation used the isolated minimax-h3-audio-T8 ComfyUI nodes with dual_clock_euler + native_flow, video shift 12, audio shift 3, and 20 joint denoising steps. That custom node package is the validation environment, not an architectural requirement of the checkpoint.

Validation

Controlled setup

ComfyUI: 0.31.0, commit cbbc9dab1f03d0d9a6caa8a8be7d77a7e37e1e44
Checkpoint family: full Ref2VA INT8 ConvRot
Turbo LoRA: none
Resolution: 512 × 512
Frames: 22 at 24 FPS
Steps: 20
Sampler/scheduler: dual_clock_euler / native_flow
Seeds: 2608103502 and 2608103503

The 22-frame test is a short diagnostic, below the official 4–15 second usage range. It does not prove quality at the recommended 124+ frame range. One centered portrait reference and one prompt were used; the reference image is not redistributed.

The original checkpoint was generated twice under seed 2608103502. Decoded video was pixel-identical and decoded PCM audio was sample-identical, confirming deterministic repeatability for this setup.

Candidate-to-source ratios

Seed	Skin gradient	Skin Laplacian	Skin high-pass	Positive skin ROIs	Temporal change	Audio correlation	Audio RMS
2608103502	1.0325	1.0760	1.0189	2/3	1.0321	0.9930	0.9189
2608103503	1.0266	1.0418	1.0131	3/3	1.0134	0.9687	1.0154

The skin high-pass proxy increased by approximately 1.89% and 1.31% across the two seeds, without a fixed 2×2 parity artifact. This is a weak proxy-level signal only. Manual inspection did not establish a convincing reduction of oily or waxy skin appearance.

The first seed also showed an approximately 8.1% audio RMS reduction. H3 predicts audio and video in a joint denoising trajectory, so changing a video-only input projection can still indirectly change the generated audio.

Full machine-readable results:

• evidence/metrics_seed_2608103502.json
• evidence/metrics_seed_2608103503.json

Visual comparisons

Each comparison is source / input HF 1.01 / input HF 1.02 from left to right. The 1.01 candidate is shown only as an additional control and is not included in this repository.

Seed 2608103502:

Seed 2608103503:

Limitations

• This is not an official MiniMax model.
• It is not a trained fine-tune, LoRA, distillation, restoration model or calibrated de-wax model.
• Only one tensor was heuristically transformed; the direction was not learned from paired data.
• Evaluation covers two seeds, one portrait reference, one prompt and a short 22-frame diagnostic.
• No official-range 124–362 frame matrix, 768p/2K test, multi-subject test, blind preference study or broad prompt evaluation has been completed.
• Small proxy-metric gains may not correspond to perceptual improvement.
• Identity, framing, motion and generated audio may change even when the modified tensor is video-specific.
• Compatibility with Turbo LoRAs, third-party samplers, alternate quantization loaders and older ComfyUI versions is not established.
• The source checkpoint remains the recommended default for production or reproducible comparisons.

Intended use

Intended only for controlled research and A/B evaluation of a small, auditable MiniMax H3 video-input projection change, subject to the upstream license and acceptable-use policy.

Not intended for claims of guaranteed quality improvement, identity preservation, medical/forensic use, deceptive impersonation, bypassing safety systems, or any use prohibited by the MiniMax H3 license.

License and attribution

MiniMax H3 and this Model Derivative are governed by the MiniMax H3 Community License Agreement. The upstream license, including territorial restrictions, redistribution requirements, commercial terms and acceptable-use policy, takes precedence over this model card.

Before distributing the checkpoint:

1. include an unmodified copy of the official MiniMax H3 LICENSE in the repository;
2. include the provided NOTICE;
3. retain a prominent statement that minimax_h3_ref2va_patchin_hf102.safetensors is modified;
4. comply with the applicable-territory and downstream-user requirements;
5. verify any commercial-use authorization requirements directly with MiniMax.

Upstream projects:

• MiniMaxAI/MiniMax-H3
• Comfy-Org/MiniMax-H3
• Official MiniMax H3 GitHub repository

Modification notice

minimax_h3_ref2va_patchin_hf102.safetensors is a modified file derived from the MiniMax H3 Ref2VA checkpoint. It was created without training by applying a 1.02 non-DC 2×2 Haar gain to video_patch_proj.weight. This modification is independent and is not endorsed by MiniMax or Comfy-Org.